Index

בס״ד

Torah

תּוֹדָהCredits and sources

Toraoke exists thanks to the work of many people and institutions: those who preserved and transcribed the text, those who translated it, those who lent their voices to the reading and the cantillation, those who built the free tools the app relies on. This page thanks them and states, for each, what we use, where it comes from and on what terms.

We chose our sources by one criterion: free access to the data. The texts, translations, analyses, recordings and tools we use are openly available, so that anyone can obtain them and check what Toraoke does with them.

Citing a source is not the same as having permission to use it. That is why next to each item we state the status of its rights as we know it today, even when the verification is not complete.

Last updated: 15 November 2026

How to read the status

  • Public domain
  • Open license
  • Declared license · authorization to be confirmed
  • Link only
  • Made for Toraoke
  • Not published

The text of the Torahתּוֹרָה

Hebrew text: Westminster Leningrad Codex

Status: Public domain

Consonants, vowels (niqqud), cantillation accents (te’amim), maqaf, paseq and the signs פ and ס of the open and closed parashot, following the Leningrad Codex (c. 1008) in the Westminster Leningrad Codex digital edition, in the version of the Open Scriptures Hebrew Bible (see below). Where the text has a ketiv and a qere, Toraoke shows the qere, that is, the form that is read.

The text of the Westminster Leningrad Codex is in the public domain.

Lemmas and grammatical analysis: Open Scriptures Hebrew Bible

Status: Open license

The lemma of each word (with its Strong’s number) and the morphological analysis from which Toraoke derives the part of speech, gender, number, person, state, conjugation and suffixes shown in the “Grammar” panel. The correspondence between the Hebrew verse numbering and that of the translations also comes from here.

Original work of the Open Scriptures Hebrew Bible available at https://github.com/openscriptures/morphhb. License CC BY 4.0; the Italian labels are Toraoke’s.

Strong’s dictionary: Open Scriptures

Status: Open license

A Concise Dictionary of the Words in the Hebrew Bible by James Strong (1894), in the Open Scriptures digital edition (electronic text by David Instone-Brewer, corrections and transliterations by David Troidl). Toraoke uses its dictionary form (lemma) of each word, in the “Grammar” panel.

Copyright 2010 Open Scriptures, Creative Commons BY-SA license (the version is not stated by the source). The original 1894 dictionary is in the public domain.

Italian translation: La Sacra Bibbia, versione Riveduta (1927)

Status: Public domain

The revision of Giovanni Diodati’s Bible carried out under the direction of Giovanni Luzzi (died 1948) and published in 1927, in the digital text distributed by eBible.org with the contribution of the Società Biblica in Italia. It is a Christian (Protestant) translation and numbers the verses as Christian Bibles do: Toraoke matches it to the Hebrew verses. In 27 verses the digital text attributed to the preceding verse the words that introduce a speech (“e disse:”, “and said:”): we have moved them into the verse they belong to in the Hebrew text.

In the public domain according to eBible.org and the Società Biblica in Italia. We are looking for an equally free Jewish Italian translation, such as the Pentateuco of Samuel David Luzzatto (1871–1876), of which no reliable digital text exists yet.

English translation: King James Version

Status: Public domain

The 1611 English translation in the standard text of 1769, in the edition with Strong’s numbers distributed by eBible.org and the CrossWire Bible Society. Those numbers allow Toraoke to link each English phrase to the Hebrew word it translates (the link is computed by Toraoke). Words in italics were added by the translators.

In the public domain outside the United Kingdom, where printing remains subject to the Crown’s letters patent.

Russian translation: Synodal

Status: Public domain

The Russian Synodal translation (Синодальный перевод), completed in 1876, from the RusSynodal module of the CrossWire Bible Society. The square brackets are those of the edition: they enclose the words added by the translators and the passages that the Synodal takes from the Septuagint and that are not in the Hebrew text.

In the public domain.

Transliteration, names of the te’amim, explanations and summaries

Status: Made for Toraoke

The Latin transliteration under the words is computed by rules written for Toraoke from the vowels, dagesh and accents. The names of the te’amim and their explanations, the grammatical labels in Italian, the summaries of the sidrot and aliyot and the links between English phrases and Hebrew words are Toraoke’s. The links between the Italian and Russian words and the Hebrew ones (the underlining while listening or when hovering with the pointer) are also computed by Toraoke, but automatically, by a statistical model and a language model (LaBSE): they are right in about nine cases out of ten and can be wrong, especially in poetic passages and where the translation is free. The division into sidrot and aliyot follows the traditional custom of the weekly reading.

External links

Status: Link only

Each verse links to its page on Sefaria and on Blue Letter Bible; in the “Grammar” panel the lemma links to the Blue Letter Bible lexicon and to the Brown-Driver-Briggs dictionary on Sefaria; the cantillation panel links to the Wikipedia articles on the te’amim.

These are only links: Toraoke does not copy the content of these pages.

Sources used in the past

Status: Link only

Until November 2026 Toraoke took the Hebrew text, the grammatical analysis, the King James Version, the Synodal and the links between English and Hebrew from the pages of Blue Letter Bible, and the Italian translation from the Bibbia CEI 2008 (© Fondazione di Religione Santi Francesco d’Assisi e Caterina da Siena) on bibbiaedu.it. We have replaced them with the free sources listed above, because those sites do not authorize copying their content.

Voices and reading traditionsקְרִיאָה

Every recording carries two sets of rights, even when the text and the traditional melodies are free: those of the person who reads or chants and those of whoever made the recording. That is why a recording can be heard on Toraoke only after its rights have been checked; being downloadable elsewhere is not enough.

Recordings you can hear on Toraoke

Rav Dan Be’eri — Reading

Status: Declared license · authorization to be confirmed

A reading of the entire Tanakh in Hebrew, without cantillation, by Rav Dan Be’eri. On Toraoke: one clip per verse, with the words highlighted while listening.

License
CC BY-SA 3.0, declared on Internet Archive by whoever uploaded the recordings
Attribution
“Reading: Rav Dan Be’eri, Tanakh Narration. Source: Internet Archive. Declared license: CC BY-SA 3.0. Audio split into verses and re-encoded.”
Changes
each chapter is split into one clip per verse and re-encoded as MP3; the start of each word is estimated automatically
Derived clips
distributed under the same license, CC BY-SA 3.0

We do not yet have written confirmation from Rav Be’eri, or from whoever holds the rights to the recordings, that they were released under this license. If that is you, or you know who it is, write to us: we will correct the credits or withdraw the recordings on request.

Rav Dan Be’eri — Cantillation

Status: Declared license · authorization to be confirmed

The Tanakh chanted by Rav Dan Be’eri with the te’amim, in the Jerusalem Sephardi tradition (sefaradi-yerushalmi), in the simplified version used in schools, as Rav Mordecai Attiya taught it (as the source describes it). On Toraoke: one clip per verse, with the words highlighted while listening.

License
CC BY-SA 3.0, declared on Internet Archive by whoever uploaded the recordings
Attribution
“Cantillation: Rav Dan Be’eri, Tanakh Cantillation (Jerusalem Sephardi tradition). Source: Internet Archive. Declared license: CC BY-SA 3.0. Audio split into verses and re-encoded; word timings estimated automatically.”
Changes
each chapter is split into one clip per verse and re-encoded as MP3; the start of each word is estimated automatically
Derived clips
distributed under the same license, CC BY-SA 3.0

We do not yet have written confirmation from Rav Be’eri, or from whoever holds the rights to the recordings, that they were released under this license. If that is you, or you know who it is, write to us: we will correct the credits or withdraw the recordings on request.

Unpublished recordings

We thank and mention them because they are part of Toraoke’s history or of the tradition we want to make known, but they cannot be heard on Toraoke: mentioning them here does not mean having permission to use them.

Moise Levy — Italian-rite cantillation

Status: Not published

The Torah chanted according to the Italian rite in the voice of Moise Levy, published by Moise Levy Editore (Milan) together with its edition of the Torah with Rashi, and available to listen to on its website, libri.levy.it.

Earlier versions of Toraoke used excerpts from these recordings. In this version they are not published: the publisher’s terms require written authorization to republish the recordings, and we do not have it. To listen to them, see the publisher’s website.

Abraham Shmuelof — Reading

Status: Not published

A reading of the Hebrew Bible (Letteris text) by Father Abraham Shmuelof (died 1994), recording ℗ 1992 Talking Bibles International, available to listen to at listen.talkingbibles.org.

Earlier versions of Toraoke used excerpts from this recording, taken from videos published on YouTube, and wrongly attributed the voice to “M.L. Kohen”, who had made those videos: we apologize. The voice is Abraham Shmuelof’s, as confirmed by the author of the videos and by the descriptions that accompany them. In this version it is not published: we do not have the producer’s permission.

Historical Italian cantillation

There has been a Jewish presence in Italy for more than two thousand years, and its communities, of the Italian, Sephardi and Ashkenazi rites, have preserved their own ways of chanting the Torah, differing from city to city. Many of these melodies survive only in field recordings and in the memory of a few readers. Making them known to those who are learning, with the respect due to those who preserved them, is one of Toraoke’s aims.

Thesaurus of Italian Jewish Music — Centro Internazionale Leo Levi

Status: Not published · under study

The Thesaurus of Italian Jewish Music of the Centro Internazionale Leo Levi in Florence collects recordings of Torah readings in many Italian communities: the field recordings made by Leo Levi in the 1950s for the Centro Nazionale Studi di Musica Popolare, now in the Archivi di Etnomusicologia of the Accademia Nazionale di Santa Cecilia; those of Giovanni Marzi in Reggio Emilia (1954); and more recent recordings. Toraoke is studying 46 of them, from 14 communities: Ancona, Casale Monferrato, Ferrara, Firenze, Livorno, Mantova, Padova, Pitigliano, Reggio Emilia, Roma, Torino, Trieste, Venezia, Verona.

None of these recordings is published on Toraoke. Before making them available to listen to, we want to clarify for each one the date, the history of its publication and the consent of the readers or their heirs, in agreement with the Centro Leo Levi. The names of the individual readers will appear here when their recordings are published with their consent.

Technology

Speech recognition: zipformer2-66m-streaming-he

Status: Open license

The karaoke recognizes the reading with the zipformer2-66m-streaming-he model by Yakov Kolani, published by ivrit.ai and trained on Hebrew recordings from ivrit.ai; it derives from a model of the icefall project (k2-fsa, Apache 2.0 license). The model (85 MB) is downloaded into the browser and recognition takes place on the device: the voice does not leave the device. The model is used as published.

License CC BY 4.0.

ONNX Runtime Web 1.23.2

Status: Open license

The engine that runs the model in the browser (WebAssembly). © Microsoft Corporation.

MIT License: the full text is in the Licenses section.

Whisper and faster-whisper

Status: Used only in preparation

The Whisper speech recognition model (OpenAI) in its “small” version, run with faster-whisper (SYSTRAN), was used during preparation to estimate the start of each word in the recordings and to check the verse cuts. It is not part of the app; the word timings it estimated are.

MIT License for both.

Typefaces

Status: Open license

Alegreya and Alegreya Sans for Latin and Cyrillic text, Noto Serif Hebrew for Hebrew (בְּרֵאשִׁית בָּרָא). They are served by this site, without external services.
Alegreya: © Huerta Tipográfica (The Alegreya Project Authors), 2011 · source · license text
Alegreya Sans: © Huerta Tipográfica (The Alegreya Sans Project Authors), 2013 · source · license text
Noto Serif Hebrew: © The Noto Project Authors, 2022 · source · license text

SIL Open Font License 1.1. The typeface names remain their authors’.

Toraoke’s code

Status: Made for Toraoke

The pages, the karaoke (audio filters, running the model, following the reading), the transliteration and the preparation tools were written for Toraoke.

Images and video

App icon

Status: Made for Toraoke

The letter א in white on a black background, designed for Toraoke with the Noto Serif Hebrew typeface.

Presentation video

Status: Provided by the project

A video in English with subtitles, edited for Toraoke from screen recordings of the app (without audio), the logo and two screenshots of public pages: Rav Be’eri’s collection on Internet Archive and the summary of the CC BY-SA 3.0 license on the Creative Commons website (Creative Commons content, CC BY 4.0 license). The animated presenter, Mitzvahboy, and his voice were provided by the project.

No third-party recording can be heard in the video.

Licenses

The licenses of the materials Toraoke publishes, with the full text where the license requires it to be reproduced. This page does not replace the original licenses, which prevail, and is not legal advice.

CC BY-SA 3.0 · legal codeRav Dan Be’eri’s recordings and the derived clips
CC BY 4.0 · legal codeLemmas and grammatical analysis (Open Scriptures Hebrew Bible); speech recognition model; Creative Commons texts in the video
SIL OFL 1.1 · legal codeAlegreya, Alegreya Sans, Noto Serif Hebrew
MITONNX Runtime Web; Whisper and faster-whisper (preparation only)
Apache 2.0icefall, from which the recognition model derives
ONNX Runtime Web license notice (MIT)
MIT License

Copyright (c) Microsoft Corporation

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

Corrections and permissions

If you hold rights to a material mentioned here, as author, performer, producer, publisher or heir, or if you find an error in a name, a date or a source, let us know: tell us the material and the page on which it appears. If a rights holder asks us to remove a material, we remove it while we verify; if they grant us its use on different terms, we update the credits as they ask.

The address for reports is in preparation and will be given here.